HEAT AND NOT LIGHT A MOTIVE TOTTED. 
133 
used. When the light of three candles passed through half an 
inch of the solution, slow rotation was produced. Two were 
blown out; the vanes stopped, and reverse motion ensued. 
The two candles were again lighted, and a short piece of 
magnesium ribbon burnt behind them. Eotation im- 
mediately quickened, and as soon as the ribbon came to an 
end, the . vanes stopped, and reverse motion ensued. The 
same effects were produced when a thickness of 8 inches of 
the solution was used, but in a less degree. I find that when- 
ever light, passing through an alum solution of whatever 
thickness, produces black repulsion, reverse motion begins as 
soon as the source of light is removed or in any way diminished. 
The more perfect the radiometer, the less this diminution need 
be. In a radiometer with bad reflecting surfaces, the black 
repulsion will only decrease in velocity. Motion may altogether 
cease if the degree of diminution of light be very great, but 
reverse motion cannot be produced. I find that the light of' 
three candles, when passed through half an inch of the alum 
solution, and causing slow black repulsion in the radiometer,, 
is capable of slightly heating the blackened globe of a dif- 
ferential thermometer, and causing a depression of the 
sulphuric acid equal to about one-thirtieth of an inch, the 
other globe being carefully sheltered from the light. I could 
not observe any depression when a greater thickness of alum 
solution was used. 
I find that a source of light, however bright, cannot produce 
rotation of the vanes of a radiometer unless it is capable of 
raising the temperature of the residual air within the globe. 
Two or three experiments, I think, prove this. I placed B 
close to a given source of bright light ; very rapid black repulsion 
at once ensued. Taking it from its stand, and heating the 
whole of the globe and stem over the flame of a spirit lamp 
until the glass was very hot, and the air inside raised to a like 
temperature, I quickly placed it in its original position, and 
found that the light had no effect, the vanes remaining per* 
fectly stationary for more than a minute ; when, the globe and 
internal air becoming cooler, black repulsion ensued. Wishing^ 
to produce this effect without sensibly heating the globe, I 
placed the stem of B in a large test tube tilled with cold 
water. The bright light of a candle placed 3 inches from 
the globe produced quick rotation. I then heated the test 
tube with the flame from a spirit lamp, taking care that the hot 
external air should not ascend and warm the globe. As the water 
became warm, and heated the air in the stem of the globe, the hot 
internal air rose, the black repulsion became less and less, and 
when the water nearly boiled all motion ceased for a few seconds, 
and reverse motion set in. The air inside now being warmer than 
